Effect of combination therapy on joint destruction in rheumatoid arthritis: a network meta-analysis of randomized controlled trials.

BACKGROUND

Despite significant cost differences, the comparative effect of combination treatments of disease modifying anti-rheumatic drugs (DMARDs) with and without biologic agents has rarely been examined. Thus we performed a network meta-analysis on the effect of combination therapies on progression of radiographic joint erosions in patients with rheumatoid arthritis (RA).

METHODS AND FINDINGS

The following combination drug therapies compared versus single DMARD were investigated: Double DMARD: 2 DMARDs (methotrexate, sulfasalazine, leflunomide, injectable gold, cyclosporine, chloroquine, azathioprin, penicillamin) or 1 DMARD plus low dose glucocorticoid (LDGC); triple DMARD: 3 DMARDs or 2 DMARDs plus LDGC; biologic combination: 1 DMARD plus biologic agent (tumor necrosis factor α inhibitor (TNFi) or abatacept or tocilizumab or CD20 inhibitor (CD20i)). Randomized controlled trials were identified in a search of electronic archives of biomedical literature and included in a star-shaped network meta-analysis and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement protocol. Effects are reported as standardized mean differences (SMD). The effects of data from 39 trials published in the period 1989-2012 were as follows: Double DMARD: -0.32 SMD (CI: -0.42, -0.22); triple DMARD: -0.46 SMD (CI: -0.60, -0.31); 1 DMARD plus TNFi: -0.30 SMD (CI: -0.36, -0.25); 1 DMARD plus abatacept: -0.20 SMD (CI: -0.33, -0.07); 1 DMARD plus tocilizumab: -0.34 SMD (CI: -0.48, -0.20); 1 DMARD plus CD20i: -0.32 SMD (CI: -0.40, -0.24). The indirect comparisons showed similar effects between combination treatments apart from triple DMARD being significantly better than abatacept plus methotrexate (-0.26 SMD (CI: -0.45, -0.07)) and TNFi plus methotrexate (-0.16 SMD (CI: -0.31, -0.01)).

CONCLUSION

Combination treatment of a biologic agent with 1 DMARD is not superior to 2-3 DMARDs including or excluding LDGC in preventing structural joint damage. Future randomized studies of biologic agents should be compared versus a combination of DMARDs.